Have a high threshold for their internal links—impossible to make causal connections with patent data

Filippo, Northwestern University Kellogg School of Management; Timothy, Boston University Questrom School of Business and NBER. “Patent policy and American innovation after eBay: An empirical examination.” Volume 48, Issue 5, June 2019, Pages 1271-1281. LJS

In economic theory, “innovation incentives” stand for a wide variety¶ of different types of risky investment. In practice, these investments are¶ notoriously difficult to measure well. Economists who study the subject¶ tend to rely on three available measures: R&D, patents and pro¶ ductivity.16 Each outcome has benefits and drawbacks. R&D investment¶ is measured in dollars– a metric that is easy to interpret– but only¶ captures inputs to the innovation process. While patents are arguably a¶ better measure of innovative output, not all inventions are patented,¶ and even when they are patented it remains hard to attach a dollar¶ value to the underlying invention. Many economists would argue that¶ the best measure of technological progress is productivity growth¶ changes in the amount of output produced for a given level of inputs.¶ However, productivity growth is difficult to measure, and often requires¶ the analyst to make strong assumptions.¶ Despite these challenges, it is still worth considering whether data¶ on R&D, patents or productivity provides any evidence that changes in¶ patent law, starting with eBay, had a measurable impact on American¶ innovation. As discussed above, our objective is not to provide a causal¶ estimate of the impact of the eBay decision. Rather, given the rhetoric¶ regarding eBay and subsequent changes in patent policy, we are asking¶ whether the available evidence can support the strong claim that eBay¶ had a large stifling impact on American innovation.¶ Ideally, we could rely on experimental evidence to evaluate claims¶ that recent changes in patent law are leading to less innovation.¶ However, a controlled experiment requires a control group, which¶ presents serious challenges when the experiment involves randomly¶ assigning current law to one set of firms (or economies) and a coun¶ terfactual legal regime to others. We consider two feasible alternatives that exploit variation in U.S. patent law over time.¶ The first alternative is to simply look for changes in the rate of in¶ novation before and after the eBay decision. While a sharp decline in¶ innovation (or its growth rate) around 2006 would be consistent with¶ the hypothesis that recent U.S. patent policy changes have been¶ harmful, we must be cautious about interpreting this type of evidence.¶ A good experiment requires that we hold “all else equal,” and in¶ practice, there are many uncontrolled differences in the pre and post¶ eBay innovation landscape. Potentially confounding factors include¶ changes in the business environment, including the rise of Non¶ Practicing Entities, and changes in the broader economic and techno¶ logical environment, such as the great recession. Nevertheless, if we fail¶ to find any evidence of an “eBay effect” in this type of pre/post analysis,¶ it would suggest that confounding factors are at least as important to¶ innovation as changes in patent policy.¶ Our second feasible alternative compares changes in innovation¶ activity around the time of the eBay ruling across groups that may have¶ been differentially affected by the decision. This type of analysis is si¶ milar to what economists call a differences-in-differences design. In this¶ type of empirical analysis, the control group serves the same purpose as¶ in the idealized experiment described above– it provides measures of¶ counterfactual outcomes (i.e. the but-for world) that we would have¶ expected for the treated. Because we could not randomly assign the¶ treatment, however, the “but for” assumption is a much stronger one in¶ our analysis below.